Consumer-driven virtual power plants: A field experiment on the adoption and use of prosocial technology

Internet connectivity and electronic innovation have opened the door for the development of demand side management (DSM) systems that seek to equalise energy supply and demand while reducing peak consumption and delivering a less carbon-intensive energy grid. Additionally, demand response strategies that rely on individual behaviour change have consistently demonstrated that flexibility in energy demand exists. By displacing demand at peak times to moments of the day during which the wind blows and the sun shines, more reliance on renewable energy can be achieved.

The authors of this paper analyse flexibility in residential energy demand through a series of experiments on the adoption and usage of WiFi-enabled smart plugs in the home. This work considers the extent to which energy consumers’ adoption of an inexpensive Internet of Things (IoT) technology, namely the smart plug, can allow for increased flexibility in energy demand, given there is an embedded stock of energy-consuming goods and durables in the economy.

The authors have piloted a novel ‘smart energy architecture’. The programme distributed smart plugs to participants living in student accommodation in the UK and remotely switched off their selected electronic or appliance randomly over a period of time. Given existing evidence of latent energy demand flexibility, the study explores the extent to which user behaviour or reservations might constrain what is technically feasible.

The authors also tested ways in which users might be incentivised to overcome such reservations by distributing monetary vouchers via a lottery system based on how much participants used the plugs. They found that the participants were more inclined to participate if the switch-off events were long enough for them to have a greater chance of winning a lottery prize. However, a higher frequency of switch-offs may reduce participation.

Understanding which information and incentives are impactful in this context is critical if IoT applications are to be effective in reducing energy use and thus cutting greenhouse gas emissions:  which ones are likely to be adopted and influence energy consumption behaviour? How do users respond to real-time feedback mechanisms?

Key points for decision-makers

• Smart plugs have the potential to be part of a smart low-carbon grid. They enable residents to automate their demand response, i.e. to switch off appliances automatically when less renewable energy is available.
• The authors study the adoption of and user interaction with devices for automated energy demand-side management. Their programme allowed them to explore whether and how devices and incentive systems may be designed and deployed to improve load balancing in a centralised energy grid – where electrical power stations store excess electrical power during low demand periods for release as demand rises.
• Through a series of randomised controlled trials in student accommodation in the UK, the authors observed how participants responded to their electricity supply being switched off, e.g. which times of day or days of the week individuals might be willing to have their appliances turned off, and when they might not be willing to do so. They would then choose to either use the ‘override’ button for the plug to ignore the random switch-off, or to unplug the appliance from the smart plug.
• The authors compared adoption and use of the smart plugs by randomly allocating participants to different treatment groups that varied both the frequency (either once a day or up to eight times a day) and duration (either 15 or 30 minutes) of switch-off events. They incentivised participation by distributing monetary vouchers via a lottery system based on how much participants used the plugs.
• Results suggest that (i) while participation is highly skewed with many users connecting very small or no power loads, as might be expected in university housing with limited appliances, there are also some ‘superusers’, connecting loads that could make a substantial contribution to load balancing when allowing them to be switched off through the smart plug; (ii) longer and less frequent switch-offs increase flexibility; and (iii) users clearly respond to reward payouts, suggesting that adoption is not purely driven by prosocial motives – e.g. the desire to reduce their greenhouse gas emissions.
• The authors posit that if scaled up, such a system would reduce the need for inefficient and often carbon-intensive back-up power generators during periods of peak demand while increasing the possible share of energy supplied by intermittent renewable energy sources.